Regression Analysis on Climate Data

1 Importing Necessary Liberaries

import pandas as pd
import plotly.express as px

2 Load the dataset

# Load the dataset
df = pd.read_csv('./processed_data/NABR_historic.csv')
df.drop(columns=['TimePeriod', 'RCP'], axis=1, inplace=True)
df.head()

	long	lat	year	scenario	treecanopy	Ann_Herb	Bare	Herb	Litter	Shrub	...	PPT_Annual	T_Winter	T_Summer	T_Annual	Tmax_Summer	Tmin_Winter	VWC_Winter_whole	VWC_Spring_whole	VWC_Summer_whole	VWC_Fall_whole
0	-110.0472	37.60413	1980	sc1	0	0	84	5	11	7	...	13.79	0.964835	23.15924	23.159240	37.05	NaN	NaN	NaN	NaN	NaN
1	-110.0472	37.60413	1980	sc1	0	0	84	5	11	7	...	2.69	0.964835	23.15924	0.964835	37.05	NaN	NaN	NaN	NaN	NaN
2	-110.0472	37.60413	1980	sc1	0	0	84	5	11	7	...	13.79	0.964835	23.15924	0.964835	37.05	NaN	NaN	NaN	NaN	NaN
3	-110.0472	37.60413	1980	sc1	0	0	84	5	11	7	...	2.69	0.964835	23.15924	23.159240	37.05	NaN	NaN	NaN	NaN	NaN
4	-110.0472	37.60413	1980	sc1	0	0	84	5	11	7	...	NaN	NaN	NaN	NaN	NaN	-12.45	0.113447	0.096831	0.041876	0.052298

5 rows × 27 columns

2.1 Missing Values and Data Types

import seaborn as sns
import matplotlib.pyplot as plt

# Visualize missing values
plt.figure(figsize=(15, 10))
sns.heatmap(df.drop(columns=['lat', 'long', 'year', 'scenario','treecanopy', 'Ann_Herb', 'Bare', 'Litter', 'Shrub'], axis=1).isnull(), cbar=False, cmap='viridis')
plt.title('Missing Data Heatmap', fontsize=16, fontweight='bold', loc='left')
plt.xticks(rotation=45) 
plt.show()

3 Data Preprocessing

# Assuming 'lat' and 'long' are the column names for latitude and longitude in your dataset
# Creating an interactive map to visualize the park location
fig = px.scatter_geo(df,
                     lat='lat',
                     lon='long',
                     scope='usa',
                     title='Utah National Parks Geographic Overview')
fig.update_geos(projection_type="natural earth")
fig.update_layout(showlegend=True)
fig.show()

4 EDA

import pandas as pd
import plotly.graph_objects as go
import plotly.offline as pyo

pyo.init_notebook_mode(connected=True)

# Assuming df is your DataFrame
# Engineering features
df['Annual_Soil_Moisture'] = df[['VWC_Winter_whole', 'VWC_Spring_whole', 'VWC_Summer_whole', 'VWC_Fall_whole']].mean(axis=1)
df['Annual_Temperature'] = df['T_Annual']
df['Annual_Precipitation'] = df['PPT_Annual']

vegetation_variables = ['treecanopy', 'Ann_Herb', 'Bare', 'Herb', 'Litter', 'Shrub']

# Set up the matplotlib figure
for veg_var in vegetation_variables:
    # Precipitation vs Vegetation
    plt.figure(figsize=(8, 5.5))
    sns.scatterplot(data=df, x='Annual_Precipitation', y=veg_var, color='#E8B989')
    plt.title(f'Annual Precipitation vs {veg_var}')
    plt.xlabel('Annual Precipitation (mm)')
    plt.ylabel(f'{veg_var} Coverage')
    plt.show()

    # Temperature vs Vegetation
    plt.figure(figsize=(8, 5.5))
    sns.scatterplot(data=df, x='Annual_Temperature', y=veg_var, color='#E8B989')
    plt.title(f'Annual Temperature vs {veg_var}')
    plt.xlabel('Annual Temperature (°C)')
    plt.ylabel(f'{veg_var} Coverage')
    plt.show()

    # Soil Moisture vs Vegetation
    plt.figure(figsize=(8, 5.5))
    sns.scatterplot(data=df, x='Annual_Soil_Moisture', y=veg_var, color='#E8B989')
    plt.title(f'Annual Soil Moisture vs {veg_var}')
    plt.xlabel('Annual Soil Moisture')
    plt.ylabel(f'{veg_var} Coverage')
    plt.show()